MATLAB: Using MATLAB in Data Science

Name: MATLAB: Using MATLAB in Data Science
Author: Gabriel Popoola

Unlock the powerful modeling capabilities of MATLAB

Preview the first lesson free — get full access to all 3 lessons.

Course: On-Demand

Welcome to the lesson on Statistical Analysis in MATLAB. During this lesson, I'm gonna go over two main categories that are important to Statistics, which are Measures of Centrality and Measures of Spread. To get started, we're gonna load a data set, and the data set that we're gonna be working with is the car data set. So to get that loaded, just run the load carbig command. And if you check your workspace, you should see a series of variables. And so, now I'm just going to go over the primary measures of centrality, which are the mean, median, and mode. The mean, which is also referred to as the arithmetic mean or the average, is a common measure of centrality. The mean is useful for symmetric distributions, but notoriously sensitive to outliers. If your data set's not distributed symmetrically, or has extreme outliers, you'll need to consider how these factors will affect the calculation of the mean. And so now in order to calculate the mean, the function is simple. We just use mean, and the variable. So, right now I'm just gonna calculate the mean of acceleration. And as you can see here, the mean is 15.5197. You can also calculate the mean of, let's say, cylinders, and you'll get 5.4754. And so, the mean by itself is an important metric or value, but it's often needed to be viewed with other measures such as the median, which gives the midpoint of the sorted data. So, half the data is greater than the median, and half the data's smaller. The median is much more resistant than the mean to changes in a few data values, and is an especially useful center for non-symmetric or skewed distributions. And to get your median, you simply use the median function, and will pass in acceleration. And you can see here, the median is 15.5. And we'll do the same for cylinders, which is four. So as you can see here, if we look at the mean of acceleration and compare it to the median of acceleration, there's only a 0.0197 difference in the two. And because the mean and median are so similar, you can assume that the distribution is nearly perfectly symmetrical. However, when we look at the mean of cylinders, we can see that the mean is 5.47, and the median is four. And so this is pretty much telling us that those distributions are not symmetrical. Next up, we have the mode, which is simply the most frequent value in an array. And if there are multiple modes, it takes the smallest. And just like the mean and median, the mode function is simple. And we'll look at the mode for acceleration, and the mode for cylinders. And we have 14.5 and four. So, this just means that a majority, or rather, the most common acceleration, is 14.5, and the most common number of cylinders is four. Next up, I'm gonna go over measures of spread, and start with the standard deviation, which like the main is typically used to measure the spread of symmetric distributions that follow a normal distribution. Because the standard deviation is the square root of the variance, the sum of the squares of the distances of data values from the mean, the standard deviation tends to amplify the effect of outliers. And so to compute the standard deviation, you use the std function, and we'll compute this for acceleration, which will give you 2.8034, and we'll compute it for cylinders, which gives us 1.7122. And the standard deviation, because they are relatively small, it shows that the data isn't widely spread in terms of distance from the mean. And the final measure that I want to discuss is the interquartile range, which is based on the median. And it gives the distance between the 25th and 75th percentile in the data. Or in other words, it gives you the width of the region that contains the middle 50% of your data values. And to compute this, we just use the iqr function, and we're going to compute it for acceleration, which gives us 3.5. And lastly, we're gonna compute it for cylinders, which gives us four. And because these values are relatively low or small, this tells you that your data doesn't deviate or spread too much from the median of the data set. This concludes this lesson. Thank you.

Intermediate Provider : Gabriel Popoola 3 Lessons · 20m in Arabic, German, English, Spanish, French, Portuguese, Chinese

Course Description

Data science is a diverse and varied field that combines elements of statistics, data visualization, predictive modeling, and data analysis. Although there are a vast number of software tools and programming languages out there that allow you to work in one or more of these aspects, there are few that allow you to work with all of them. This is what makes MATLAB such a powerful programming language and IDE, as it offers an extensive set of features and capabilities.

In this module we take a look at the powerful modeling capabilities of MATLAB, in which you will learn about data visualization in MATLAB using the visualization toolbox, and how to generate different, legible figures representing your data and insights.

These figures include line, scatter, histogram, and area plots. You will also learn the difference between selection, plots, and options in MATLAB, as well as how to load built-in data sets.

What You'll Learn

Understand the different categories of statistical analysis
Assess the usefulness of various predictive models
Set up your data so it can be properly used in MATLAB apps for model selection
Visualize data in MATLAB using the visualization toolbox to generate legible figures
Create line, scatter, histogram, and area plots to represent data and insights
Distinguish between selection, plots, and options in MATLAB and load built-in data sets

Key Takeaways

MATLAB is a programming language and IDE that supports statistics, data visualization, predictive modeling, and data analysis together within one tool.
The visualization toolbox in MATLAB can generate different legible figures, including line, scatter, histogram, and area plots.
The course covers the difference between selection, plots, and options in MATLAB, as well as how to load built-in data sets.
Data must be set up properly so that it can be used in MATLAB apps for model selection.

Frequently Asked Questions

What will I learn in this MATLAB data science course?

You will learn about data visualization in MATLAB using the visualization toolbox, how to generate figures such as line, scatter, histogram, and area plots, the categories of statistical analysis, how to assess predictive models, and how to set up data for model selection in MATLAB apps.

What topics does this course cover?

The course covers three lessons: Statistical Analysis, Predictive Modeling, and Model Selection.

What skills does this course help develop?

It addresses skills including Data Science, MATLAB, Statistical Classification, Statistical Modeling, Statistical Programming, and SAS Certified Data Scientist.

Why use MATLAB for data science?

Data science combines statistics, data visualization, predictive modeling, and data analysis, and MATLAB is a powerful language and IDE because it offers an extensive set of features that allow you to work with all of these aspects.

What types of plots can I create in MATLAB after this course?

You will learn to generate line, scatter, histogram, and area plots to represent your data and insights.